#Import Libraries:

import numpy as np              # for data processing.
import pandas as pd             # is an extended version of "np".

import matplotlib.pyplot as plt # for data visualization.
import seaborn as sns           #  is an extended version of "plt".

import plotly.express as px     # for more appealing interactive data visualizations.

import warnings                 # for ignoring warnings.
warnings.filterwarnings('ignore')

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC
from imblearn.over_sampling import ADASYN
from imblearn.under_sampling import TomekLinks
from sklearn.gaussian_process import GaussianProcessClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.neural_network import MLPClassifier
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
from sklearn.linear_model import SGDClassifier
from minepy import MINE

df = pd.read_csv("dataset/Employee_Attrition.csv")
df.head(10).style.background_gradient(cmap="Reds")

c=df.corr(numeric_only=True)[["DailyRate","Age"]]
c

plt.figure(figsize=(10,10))
sns.heatmap(c,annot=True,fmt=".2",cmap="Reds")
plt.title("correlation between DailyRate and Age",c="k",fontsize=17)
plt.show()

sns.countplot(x="Education",data=df)
plt.title("Count Of Education Number",c="k",fontsize=17)
plt.xlabel("Education",c="g",fontsize=16)
plt.ylabel("Count",c="r",fontsize=16)
plt.show()

sns.countplot(x="EducationField",data=df)
plt.title("Count Of EducationField Number",c="k",fontsize=17)
plt.xlabel("EducationField",c="g",fontsize=16)
plt.ylabel("count",c="r",fontsize=16)
plt.show()

df.hist(bins=30,figsize=(16,16))
plt.show()

plt.figure(figsize=(15,15))
sns.heatmap(df.corr(numeric_only=True),annot=True,fmt=".1")
plt.title("Correlation Between Columns",c="k",fontsize=25)
plt.show()

sns.catplot(x="Education",y="Age",data=df,kind="box")
plt.title("Boxplot For Education",c="k",fontsize=20)
plt.xlabel("Education",c="r",fontsize=16)
plt.ylabel("Age",c="g",fontsize=16)
plt.show()

sns.pairplot(df,vars=["Age","DailyRate"])
plt.show()

plt.scatter(x=df["TotalWorkingYears"],y=df["DailyRate"],c="r")
plt.title("Distribution 2 Columns by Scatter",c="k",fontsize=20)
plt.xlabel("TotalWorkingYears",c="y",fontsize=16)
plt.ylabel("DailyRate",c="g",fontsize=16)

Text(0, 0.5, 'DailyRate')

plt.scatter(x=df["Education"],y=df["Age"],c="b")
plt.title("Distribution 2 Columns by Scatter",c="k",fontsize=20)
plt.xlabel("Education",c="y",fontsize=16)
plt.ylabel("Age",c="g",fontsize=16)

Text(0, 0.5, 'Age')

plt.figure(figsize=(8,5))
df["Age"].plot()
df["Education"].plot()
plt.title("Distribution Age and Education",c="k",fontsize=20)
plt.xlabel("None",fontsize=16,c="r")
plt.ylabel("None",fontsize=16,c="g")
plt.legend(["Age","DailyRate"])
plt.tight_layout()
plt.show()

sns.barplot(data=df,x="EducationField",y="Education")
plt.title("Distribution of barplot",c="k",fontsize=20)
plt.xlabel("EducationField",c="r",fontsize=16)
plt.ylabel("Education",c="g",fontsize=16)
plt.show()

plt.figure(figsize=(7,7))
df.Attrition.value_counts().plot(kind="bar")
plt.show()

df.groupby("EmployeeNumber")["EmployeeCount"].mean().sort_values().head(8).plot(kind="bar")
plt.title("Distribution of groupby EmployeeCount and EmployeeNumber",c="k",fontsize=20)
plt.xlabel("EmployeeNumber",c="r",fontsize=16)
plt.show()

plt.figure(figsize=(11,11))
columns=df.columns[23:34]
for i,column in enumerate(columns,start=1):
    plt.subplot(5,4,i)
    sns.boxplot(x=df[column],color="y")
    
plt.tight_layout()
plt.show()

plt.figure(figsize=(15,15))
columns=df.columns[23:30]
for i,column in enumerate(columns,start=1):
    plt.subplot(7,1,i)
    sns.distplot(x=df[column],color="b")
    
plt.tight_layout()
plt.show()

fig=px.histogram(df,x="Education",color="Education")
fig.show()

fig=px.histogram(df,x="Attrition",color="Attrition")
fig.show()

fig=px.histogram(df,x="EducationField",color="EducationField")
fig.show()

sns.catplot(x="Attrition",y="Age",data=df,kind="box")
plt.title("Boxplot For Attrition",c="k",fontsize=20)
plt.xlabel("Attrition",c="r",fontsize=16)
plt.ylabel("Age",c="g",fontsize=16)
plt.show()

b=df.boxplot(column="Age",by="Education",figsize=(10,6))
plt.title("Plotting boxplot to check outliers",fontsize=20,c="k")
plt.xlabel("Education",fontsize=16,c="g")
plt.ylabel("outliers",fontsize=16,c="r")
b

<Axes: title={'center': 'Plotting boxplot to check outliers'}, xlabel='Education', ylabel='outliers'>

sns.pairplot(df,vars=["Education","EducationField"])
plt.show()

plt.figure(figsize=(12,5))
df["Age"].plot(c="r")
plt.title("Distribution Age",c="k",fontsize=20)
plt.xlabel("None",fontsize=16,c="r")
plt.ylabel("None",fontsize=16,c="g")
plt.legend(["Age"])
plt.tight_layout()
plt.show()

sns.catplot(x="Education",
            y="EducationField",
            hue="Attrition",
            data=df,
            kind="bar")
plt.show()

sns.jointplot(x="Age",y="DailyRate",data=df)
plt.show()

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1470 entries, 0 to 1469
Data columns (total 35 columns):
 #   Column                    Non-Null Count  Dtype 
---  ------                    --------------  ----- 
 0   Age                       1470 non-null   int64 
 1   Attrition                 1470 non-null   object
 2   BusinessTravel            1470 non-null   object
 3   DailyRate                 1470 non-null   int64 
 4   Department                1470 non-null   object
 5   DistanceFromHome          1470 non-null   int64 
 6   Education                 1470 non-null   int64 
 7   EducationField            1470 non-null   object
 8   EmployeeCount             1470 non-null   int64 
 9   EmployeeNumber            1470 non-null   int64 
 10  EnvironmentSatisfaction   1470 non-null   int64 
 11  Gender                    1470 non-null   object
 12  HourlyRate                1470 non-null   int64 
 13  JobInvolvement            1470 non-null   int64 
 14  JobLevel                  1470 non-null   int64 
 15  JobRole                   1470 non-null   object
 16  JobSatisfaction           1470 non-null   int64 
 17  MaritalStatus             1470 non-null   object
 18  MonthlyIncome             1470 non-null   int64 
 19  MonthlyRate               1470 non-null   int64 
 20  NumCompaniesWorked        1470 non-null   int64 
 21  Over18                    1470 non-null   object
 22  OverTime                  1470 non-null   object
 23  PercentSalaryHike         1470 non-null   int64 
 24  PerformanceRating         1470 non-null   int64 
 25  RelationshipSatisfaction  1470 non-null   int64 
 26  StandardHours             1470 non-null   int64 
 27  StockOptionLevel          1470 non-null   int64 
 28  TotalWorkingYears         1470 non-null   int64 
 29  TrainingTimesLastYear     1470 non-null   int64 
 30  WorkLifeBalance           1470 non-null   int64 
 31  YearsAtCompany            1470 non-null   int64 
 32  YearsInCurrentRole        1470 non-null   int64 
 33  YearsSinceLastPromotion   1470 non-null   int64 
 34  YearsWithCurrManager      1470 non-null   int64 
dtypes: int64(26), object(9)
memory usage: 402.1+ KB

df.shape

(1470, 35)

df.isna().sum().sum()

0

df.describe()

categoriacal_cols = []
for cols in df.columns:
    if(df[cols].dtype == 'object'):
        categoriacal_cols.append(cols)
print(categoriacal_cols)
print(len(categoriacal_cols))

['Attrition', 'BusinessTravel', 'Department', 'EducationField', 'Gender', 'JobRole', 'MaritalStatus', 'Over18', 'OverTime']
9

#Unique Values in each categorical columns in the dataset
unique_counts = df[categoriacal_cols].nunique()
print("Number of unique values in each categorical column:")
print(unique_counts)

Number of unique values in each categorical column:
Attrition         2
BusinessTravel    3
Department        3
EducationField    6
Gender            2
JobRole           9
MaritalStatus     3
Over18            1
OverTime          2
dtype: int64

from sklearn.preprocessing import LabelEncoder

# Create a copy of the DataFrame to avoid modifying the original one
df_label_encoded = df.copy()

# Initialize the LabelEncoder
label_encoder = LabelEncoder()

# Apply LabelEncoder to each categorical column
for col in categoriacal_cols:
    df_label_encoded[col] = label_encoder.fit_transform(df_label_encoded[col])

df_label_encoded.head()

## split data to x , y

x=df.drop("Attrition",axis=1)
y=df["Attrition"]

y.value_counts()

Attrition
No     1233
Yes     237
Name: count, dtype: int64

df['Attrition'].value_counts().plot(kind='barh', figsize=(8,6))
plt.xlabel("Count", labelpad=15)
plt.ylabel("Attrition", labelpad=15)
plt.title("Count of the Y")

Text(0.5, 1.0, 'Count of the Y')

# List of columns with more than 50% missing values
def columns_with_missing_data(df):
    missing_data_threshold = len(df) * 0.5  # 50% threshold
    missing_columns = []
    for column in df.columns:
        if df[column].isnull().sum() > missing_data_threshold:
            missing_columns.append(column)
    return missing_columns

missing_columns = columns_with_missing_data(df)
print("Columns with more than 50% missing values:")
print(missing_columns)

Columns with more than 50% missing values:
[]

from ydata_profiling import ProfileReport
ProfileReport(df)

Summarize dataset:   0%|          | 0/5 [00:00<?, ?it/s]

Generate report structure:   0%|          | 0/1 [00:00<?, ?it/s]

Render HTML:   0%|          | 0/1 [00:00<?, ?it/s]

df_label_encoded.shape

(1470, 35)

print("Intial Y value count:",y.value_counts())

Intial Y value count: Attrition
No     1233
Yes     237
Name: count, dtype: int64

X_encoded = df_label_encoded.drop("Attrition",axis=1)
y_encoded = df_label_encoded["Attrition"]

# Initialize ADASYN for oversampling
adasyn = ADASYN()
# Initialize TomekLinks for undersampling
tomek_links = TomekLinks()

# Oversample the DataFrame
X_resampled, y_resampled = adasyn.fit_resample(X_encoded, y_encoded) 

# Undersample the oversampled data
X_resampled_tomek, y_resampled_tomek = tomek_links.fit_resample(X_resampled, y_resampled)

# 1. MIC Feature Selection

def calculate_mic(X, y):
    mine = MINE()
    mic_scores = []
    for column in X.columns:
        mine.compute_score(X[column], y)
        mic_scores.append(mine.mic())
    return mic_scores

from minepy import MINE

# Initialize the MINE object
mine = MINE()

# Calculate MIC scores
mic_scores = []
for col in X_resampled.columns:
    mine.compute_score(X_resampled[col], y_resampled)
    mic_scores.append((col, mine.mic()))

# Create a DataFrame with MIC scores and sort by score
mic_df = pd.DataFrame(mic_scores, columns=['Feature', 'MIC Score'])
mic_df = mic_df.sort_values(by='MIC Score', ascending=False).reset_index(drop=True)
mic_df['Rank'] = mic_df.index + 1

print(mic_df)

                     Feature  MIC Score  Rank
0              MonthlyIncome   0.171626     1
1                MonthlyRate   0.127803     2
2                  DailyRate   0.107833     3
3             EmployeeNumber   0.103291     4
4           StockOptionLevel   0.102276     5
5       YearsWithCurrManager   0.093119     6
6             JobInvolvement   0.089979     7
7         YearsInCurrentRole   0.083455     8
8            JobSatisfaction   0.081255     9
9          TotalWorkingYears   0.080156    10
10            YearsAtCompany   0.079205    11
11                  JobLevel   0.074777    12
12   EnvironmentSatisfaction   0.073540    13
13            BusinessTravel   0.069364    14
14                       Age   0.060140    15
15                HourlyRate   0.051486    16
16          DistanceFromHome   0.050656    17
17           WorkLifeBalance   0.049084    18
18     TrainingTimesLastYear   0.045965    19
19                 Education   0.045274    20
20  RelationshipSatisfaction   0.044554    21
21            EducationField   0.042495    22
22         PerformanceRating   0.024657    23
23                   JobRole   0.024314    24
24                    Gender   0.022219    25
25         PercentSalaryHike   0.019369    26
26   YearsSinceLastPromotion   0.018624    27
27        NumCompaniesWorked   0.017293    28
28                Department   0.010029    29
29                  OverTime   0.008865    30
30             MaritalStatus   0.003764    31
31             StandardHours   0.000000    32
32             EmployeeCount   0.000000    33
33                    Over18   0.000000    34

# Select top features 
top_features = mic_df['Feature'][:25].tolist()
X_selected = X_resampled_tomek[top_features]

X_selected

# Split the data
X_train, X_test, y_train, y_test = train_test_split(X_selected, y_resampled_tomek, test_size=0.2, random_state=42)

# Scale the features
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Conv1D, MaxPooling1D, Flatten, LSTM, Bidirectional

# If GPU is available, set memory growth
gpus = tf.config.experimental.list_physical_devices('GPU')
if gpus:
    try:
        for gpu in gpus:
            tf.config.experimental.set_memory_growth(gpu, True)
    except RuntimeError as e:
        print(e)

# Function to create and train models
def create_and_train_model(model_func, X_train, y_train, model_name):
    with tf.device('/GPU:0'):
        model = model_func(X_train.shape[1])
        history = model.fit(
            X_train, y_train, 
            epochs=10, 
            batch_size=32, 
            validation_split=0.2,
            verbose=1
        )
    print(f"{model_name} training completed.")
    return model, history

# 1. Multi-layer Perceptron (MLP)

def create_mlp_model(input_shape):
    model = Sequential([
        Dense(64, activation='relu', input_shape=(input_shape,)),
        Dense(32, activation='relu'),
        Dense(16, activation='relu'),
        Dense(1, activation='sigmoid')
    ])
    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
    return model

mlp_model, mlp_history = create_and_train_model(create_mlp_model, X_train_scaled, y_train, "MLP")

Epoch 1/10
47/47 [==============================] - 2s 11ms/step - loss: 0.6034 - accuracy: 0.6806 - val_loss: 0.5163 - val_accuracy: 0.7473
Epoch 2/10
47/47 [==============================] - 0s 4ms/step - loss: 0.4487 - accuracy: 0.8011 - val_loss: 0.4204 - val_accuracy: 0.8032
Epoch 3/10
47/47 [==============================] - 0s 5ms/step - loss: 0.3793 - accuracy: 0.8323 - val_loss: 0.4010 - val_accuracy: 0.8245
Epoch 4/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3459 - accuracy: 0.8596 - val_loss: 0.3932 - val_accuracy: 0.8271
Epoch 5/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3209 - accuracy: 0.8643 - val_loss: 0.3731 - val_accuracy: 0.8404
Epoch 6/10
47/47 [==============================] - 0s 4ms/step - loss: 0.2959 - accuracy: 0.8816 - val_loss: 0.3652 - val_accuracy: 0.8404
Epoch 7/10
47/47 [==============================] - 0s 5ms/step - loss: 0.2723 - accuracy: 0.8916 - val_loss: 0.3661 - val_accuracy: 0.8298
Epoch 8/10
47/47 [==============================] - 0s 4ms/step - loss: 0.2532 - accuracy: 0.9009 - val_loss: 0.3608 - val_accuracy: 0.8457
Epoch 9/10
47/47 [==============================] - 0s 4ms/step - loss: 0.2324 - accuracy: 0.9108 - val_loss: 0.3481 - val_accuracy: 0.8431
Epoch 10/10
47/47 [==============================] - 0s 4ms/step - loss: 0.2144 - accuracy: 0.9188 - val_loss: 0.3492 - val_accuracy: 0.8484
MLP training completed.

# 2. Convolutional Neural Network (CNN)
def create_cnn_model(input_shape):
    model = Sequential([
        Conv1D(filters=64, kernel_size=3, activation='relu', input_shape=(input_shape, 1)),
        MaxPooling1D(pool_size=2),
        Flatten(),
        Dense(50, activation='relu'),
        Dense(1, activation='sigmoid')
    ])
    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
    return model

X_train_cnn = X_train_scaled.reshape((X_train_scaled.shape[0], X_train_scaled.shape[1], 1))
cnn_model, cnn_history = create_and_train_model(create_cnn_model, X_train_cnn, y_train, "CNN")

Epoch 1/10
47/47 [==============================] - 3s 9ms/step - loss: 0.5107 - accuracy: 0.7545 - val_loss: 0.4677 - val_accuracy: 0.8059
Epoch 2/10
47/47 [==============================] - 0s 5ms/step - loss: 0.4028 - accuracy: 0.8197 - val_loss: 0.4306 - val_accuracy: 0.8085
Epoch 3/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3830 - accuracy: 0.8283 - val_loss: 0.4323 - val_accuracy: 0.8005
Epoch 4/10
47/47 [==============================] - 0s 5ms/step - loss: 0.3655 - accuracy: 0.8357 - val_loss: 0.4131 - val_accuracy: 0.8165
Epoch 5/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3549 - accuracy: 0.8443 - val_loss: 0.4133 - val_accuracy: 0.8138
Epoch 6/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3490 - accuracy: 0.8470 - val_loss: 0.4245 - val_accuracy: 0.8138
Epoch 7/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3390 - accuracy: 0.8456 - val_loss: 0.4090 - val_accuracy: 0.8165
Epoch 8/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3226 - accuracy: 0.8643 - val_loss: 0.4030 - val_accuracy: 0.8191
Epoch 9/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3238 - accuracy: 0.8623 - val_loss: 0.4043 - val_accuracy: 0.8218
Epoch 10/10
47/47 [==============================] - 0s 4ms/step - loss: 0.3206 - accuracy: 0.8563 - val_loss: 0.4002 - val_accuracy: 0.8245
CNN training completed.

# 3. Recurrent Neural Network (RNN)
def create_rnn_model(input_shape):
    model = Sequential([
        LSTM(50, activation='relu', input_shape=(input_shape, 1)),
        Dense(1, activation='sigmoid')
    ])
    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
    return model

rnn_model, rnn_history = create_and_train_model(create_rnn_model, X_train_cnn, y_train, "RNN")

WARNING:tensorflow:Layer lstm will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.
Epoch 1/10
47/47 [==============================] - 3s 44ms/step - loss: 0.6442 - accuracy: 0.6919 - val_loss: 0.5510 - val_accuracy: 0.7048
Epoch 2/10
47/47 [==============================] - 2s 39ms/step - loss: 0.5381 - accuracy: 0.7532 - val_loss: 0.5395 - val_accuracy: 0.7261
Epoch 3/10
47/47 [==============================] - 2s 38ms/step - loss: 0.5181 - accuracy: 0.7585 - val_loss: 0.5232 - val_accuracy: 0.7473
Epoch 4/10
47/47 [==============================] - 2s 38ms/step - loss: 0.5086 - accuracy: 0.7565 - val_loss: 0.5131 - val_accuracy: 0.7580
Epoch 5/10
47/47 [==============================] - 2s 41ms/step - loss: 0.5096 - accuracy: 0.7438 - val_loss: 0.5090 - val_accuracy: 0.7447
Epoch 6/10
47/47 [==============================] - 2s 38ms/step - loss: 0.5017 - accuracy: 0.7658 - val_loss: 0.5120 - val_accuracy: 0.7447
Epoch 7/10
47/47 [==============================] - 2s 40ms/step - loss: 0.4980 - accuracy: 0.7565 - val_loss: 0.5036 - val_accuracy: 0.7367
Epoch 8/10
47/47 [==============================] - 2s 44ms/step - loss: 0.4908 - accuracy: 0.7532 - val_loss: 0.5428 - val_accuracy: 0.6995
Epoch 9/10
47/47 [==============================] - 2s 43ms/step - loss: 0.4973 - accuracy: 0.7611 - val_loss: 0.4933 - val_accuracy: 0.7686
Epoch 10/10
47/47 [==============================] - 2s 37ms/step - loss: 0.4914 - accuracy: 0.7651 - val_loss: 0.5150 - val_accuracy: 0.7420
RNN training completed.

# 4. CNN-BLSTM
def create_cnn_blstm_model(input_shape):
    model = Sequential([
        Conv1D(filters=64, kernel_size=3, activation='relu', input_shape=(input_shape, 1)),
        MaxPooling1D(pool_size=2),
        Bidirectional(LSTM(50, return_sequences=True)),
        Bidirectional(LSTM(25)),
        Dense(1, activation='sigmoid')
    ])
    model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
    return model

cnn_blstm_model, cnn_blstm_history = create_and_train_model(create_cnn_blstm_model, X_train_cnn, y_train, "CNN-BLSTM")

Epoch 1/10
47/47 [==============================] - 6s 37ms/step - loss: 0.5642 - accuracy: 0.7099 - val_loss: 0.5014 - val_accuracy: 0.7606
Epoch 2/10
47/47 [==============================] - 1s 14ms/step - loss: 0.5029 - accuracy: 0.7631 - val_loss: 0.4964 - val_accuracy: 0.7633
Epoch 3/10
47/47 [==============================] - 1s 14ms/step - loss: 0.4964 - accuracy: 0.7685 - val_loss: 0.4860 - val_accuracy: 0.7606
Epoch 4/10
47/47 [==============================] - 1s 13ms/step - loss: 0.4897 - accuracy: 0.7718 - val_loss: 0.4835 - val_accuracy: 0.7686
Epoch 5/10
47/47 [==============================] - 1s 13ms/step - loss: 0.4783 - accuracy: 0.7791 - val_loss: 0.4776 - val_accuracy: 0.7686
Epoch 6/10
47/47 [==============================] - 1s 15ms/step - loss: 0.4822 - accuracy: 0.7691 - val_loss: 0.4754 - val_accuracy: 0.7713
Epoch 7/10
47/47 [==============================] - 1s 13ms/step - loss: 0.4704 - accuracy: 0.7884 - val_loss: 0.4708 - val_accuracy: 0.7739
Epoch 8/10
47/47 [==============================] - 1s 13ms/step - loss: 0.4690 - accuracy: 0.7844 - val_loss: 0.4629 - val_accuracy: 0.7660
Epoch 9/10
47/47 [==============================] - 1s 13ms/step - loss: 0.4574 - accuracy: 0.7971 - val_loss: 0.4557 - val_accuracy: 0.7660
Epoch 10/10
47/47 [==============================] - 1s 14ms/step - loss: 0.4486 - accuracy: 0.7937 - val_loss: 0.4723 - val_accuracy: 0.7686
CNN-BLSTM training completed.

import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score


# Evaluate models
def evaluate_model(model, X_test, y_test, model_name):
    # Make predictions
    y_pred = model.predict(X_test)
    y_pred_classes = (y_pred > 0.5).astype(int).flatten()
    
    # Calculate metrics
    accuracy = accuracy_score(y_test, y_pred_classes)
    precision = precision_score(y_test, y_pred_classes, average='weighted')
    recall = recall_score(y_test, y_pred_classes, average='weighted')
    f1 = f1_score(y_test, y_pred_classes, average='weighted')
    
    print(f"{model_name} - Test accuracy: {accuracy:.4f}, Precision: {precision:.4f}, Recall: {recall:.4f}, F1 Score: {f1:.4f}")
    
    return accuracy, precision, recall, f1

# Evaluate each model
mlp_metrics = evaluate_model(mlp_model, X_test_scaled, y_test, "MLP")
cnn_metrics = evaluate_model(cnn_model, X_test_scaled, y_test, "CNN")
rnn_metrics = evaluate_model(rnn_model, X_test_scaled, y_test, "RNN")
cnn_blstm_metrics = evaluate_model(cnn_blstm_model, X_test_scaled, y_test, "CNN-BLSTM")

15/15 [==============================] - 0s 1ms/step
MLP - Test accuracy: 0.8362, Precision: 0.8371, Recall: 0.8362, F1 Score: 0.8358
15/15 [==============================] - 0s 2ms/step
CNN - Test accuracy: 0.8213, Precision: 0.8251, Recall: 0.8213, F1 Score: 0.8202
15/15 [==============================] - 0s 5ms/step
RNN - Test accuracy: 0.7681, Precision: 0.7867, Recall: 0.7681, F1 Score: 0.7626
15/15 [==============================] - 1s 6ms/step
CNN-BLSTM - Test accuracy: 0.7681, Precision: 0.7829, Recall: 0.7681, F1 Score: 0.7635

# Create a bar chart to compare metrics
models = ['MLP', 'CNN', 'RNN', 'CNN-BLSTM']
metrics = ['Accuracy', 'Precision', 'Recall', 'F1 Score']

fig, ax = plt.subplots(figsize=(12, 6))

x = np.arange(len(models))
width = 0.2

for i, metric in enumerate(metrics):
    values = [mlp_metrics[i], cnn_metrics[i], rnn_metrics[i], cnn_blstm_metrics[i]]
    ax.bar(x + i*width, values, width, label=metric)

ax.set_ylabel('Scores')
ax.set_title('Model Performance Comparison')
ax.set_xticks(x + width * 1.5)
ax.set_xticklabels(models)
ax.legend(loc='lower right')
ax.set_ylim(0, 1)

plt.tight_layout()
plt.show()

# Create a radar chart to compare models
fig, ax = plt.subplots(figsize=(10, 10), subplot_kw=dict(projection='polar'))

angles = np.linspace(0, 2*np.pi, len(metrics), endpoint=False)
angles = np.concatenate((angles, [angles[0]]))

for model, model_metrics in zip(models, [mlp_metrics, cnn_metrics, rnn_metrics, cnn_blstm_metrics]):
    values = list(model_metrics) + [model_metrics[0]]
    ax.plot(angles, values, 'o-', linewidth=2, label=model)
    ax.fill(angles, values, alpha=0.25)

ax.set_thetagrids(angles[:-1] * 180/np.pi, metrics)
ax.set_ylim(0, 1)
ax.set_title('Model Performance Comparison')
ax.legend(loc='upper right', bbox_to_anchor=(1.3, 1.0))

plt.tight_layout()
plt.show()

	DailyRate	Age
Age	0.010661	1.000000
DailyRate	1.000000	0.010661
DistanceFromHome	-0.004985	-0.001686
Education	-0.016806	0.208034
EmployeeCount	NaN	NaN
EmployeeNumber	-0.050990	-0.010145
EnvironmentSatisfaction	0.018355	0.010146
HourlyRate	0.023381	0.024287
JobInvolvement	0.046135	0.029820
JobLevel	0.002966	0.509604
JobSatisfaction	0.030571	-0.004892
MonthlyIncome	0.007707	0.497855
MonthlyRate	-0.032182	0.028051
NumCompaniesWorked	0.038153	0.299635
PercentSalaryHike	0.022704	0.003634
PerformanceRating	0.000473	0.001904
RelationshipSatisfaction	0.007846	0.053535
StandardHours	NaN	NaN
StockOptionLevel	0.042143	0.037510
TotalWorkingYears	0.014515	0.680381
TrainingTimesLastYear	0.002453	-0.019621
WorkLifeBalance	-0.037848	-0.021490
YearsAtCompany	-0.034055	0.311309
YearsInCurrentRole	0.009932	0.212901
YearsSinceLastPromotion	-0.033229	0.216513
YearsWithCurrManager	-0.026363	0.202089

	Age	DailyRate	DistanceFromHome	Education	EmployeeCount	EmployeeNumber	EnvironmentSatisfaction	HourlyRate	JobInvolvement	JobLevel	...	RelationshipSatisfaction	StandardHours	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	YearsAtCompany	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager
count	1470.000000	1470.000000	1470.000000	1470.000000	1470.0	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	...	1470.000000	1470.0	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000	1470.000000
mean	36.923810	802.485714	9.192517	2.912925	1.0	1024.865306	2.721769	65.891156	2.729932	2.063946	...	2.712245	80.0	0.793878	11.279592	2.799320	2.761224	7.008163	4.229252	2.187755	4.123129
std	9.135373	403.509100	8.106864	1.024165	0.0	602.024335	1.093082	20.329428	0.711561	1.106940	...	1.081209	0.0	0.852077	7.780782	1.289271	0.706476	6.126525	3.623137	3.222430	3.568136
min	18.000000	102.000000	1.000000	1.000000	1.0	1.000000	1.000000	30.000000	1.000000	1.000000	...	1.000000	80.0	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000
25%	30.000000	465.000000	2.000000	2.000000	1.0	491.250000	2.000000	48.000000	2.000000	1.000000	...	2.000000	80.0	0.000000	6.000000	2.000000	2.000000	3.000000	2.000000	0.000000	2.000000
50%	36.000000	802.000000	7.000000	3.000000	1.0	1020.500000	3.000000	66.000000	3.000000	2.000000	...	3.000000	80.0	1.000000	10.000000	3.000000	3.000000	5.000000	3.000000	1.000000	3.000000
75%	43.000000	1157.000000	14.000000	4.000000	1.0	1555.750000	4.000000	83.750000	3.000000	3.000000	...	4.000000	80.0	1.000000	15.000000	3.000000	3.000000	9.000000	7.000000	3.000000	7.000000
max	60.000000	1499.000000	29.000000	5.000000	1.0	2068.000000	4.000000	100.000000	4.000000	5.000000	...	4.000000	80.0	3.000000	40.000000	6.000000	4.000000	40.000000	18.000000	15.000000	17.000000

	MonthlyIncome	MonthlyRate	DailyRate	EmployeeNumber	StockOptionLevel	YearsWithCurrManager	JobInvolvement	YearsInCurrentRole	JobSatisfaction	TotalWorkingYears	...	HourlyRate	DistanceFromHome	WorkLifeBalance	TrainingTimesLastYear	Education	RelationshipSatisfaction	EducationField	PerformanceRating	JobRole	Gender
0	5993	19479	1102	1	0	5	3	4	4	8	...	94	1	1	0	2	1	1	3	7	0
1	5130	24907	279	2	1	7	2	7	2	10	...	61	8	3	3	1	4	1	4	6	1
2	2090	2396	1373	4	0	0	2	0	3	7	...	92	2	3	3	2	2	4	3	2	1
3	2909	23159	1392	5	0	0	3	7	3	8	...	56	3	3	3	4	3	1	3	6	0
4	3468	16632	591	7	1	2	3	2	2	6	...	40	2	3	3	1	4	3	3	2	1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2344	6734	12173	962	1737	0	0	1	0	2	10	...	34	1	3	2	2	2	1	3	7	1
2345	10761	15955	491	1643	0	2	2	1	1	18	...	41	17	2	2	3	2	1	3	7	1
2346	10736	15791	512	1536	0	3	2	1	1	18	...	42	15	2	2	3	2	1	3	7	1
2347	10784	16387	429	2006	1	0	2	2	1	21	...	42	25	3	3	3	2	2	3	6	0
2348	11186	16654	561	1903	0	0	2	1	1	20	...	47	27	2	2	3	2	1	3	7	0

Data Preprocessing

Oversampling and Undersampling

MIC Feature Selection

Model Training :

	Age	Attrition	BusinessTravel	DailyRate	Department	DistanceFromHome	Education	EducationField	EmployeeCount	EmployeeNumber	EnvironmentSatisfaction	Gender	HourlyRate	JobInvolvement	JobLevel	JobRole	JobSatisfaction	MaritalStatus	MonthlyIncome	MonthlyRate	NumCompaniesWorked	Over18	OverTime	PercentSalaryHike	PerformanceRating	RelationshipSatisfaction	StandardHours	StockOptionLevel	TotalWorkingYears	TrainingTimesLastYear	WorkLifeBalance	YearsAtCompany	YearsInCurrentRole	YearsSinceLastPromotion	YearsWithCurrManager
0	41	Yes	Travel_Rarely	1102	Sales	1	2	Life Sciences	1	1	2	Female	94	3	2	Sales Executive	4	Single	5993	19479	8	Y	Yes	11	3	1	80	0	8	0	1	6	4	0	5
1	49	No	Travel_Frequently	279	Research & Development	8	1	Life Sciences	1	2	3	Male	61	2	2	Research Scientist	2	Married	5130	24907	1	Y	No	23	4	4	80	1	10	3	3	10	7	1	7
2	37	Yes	Travel_Rarely	1373	Research & Development	2	2	Other	1	4	4	Male	92	2	1	Laboratory Technician	3	Single	2090	2396	6	Y	Yes	15	3	2	80	0	7	3	3	0	0	0	0
3	33	No	Travel_Frequently	1392	Research & Development	3	4	Life Sciences	1	5	4	Female	56	3	1	Research Scientist	3	Married	2909	23159	1	Y	Yes	11	3	3	80	0	8	3	3	8	7	3	0
4	27	No	Travel_Rarely	591	Research & Development	2	1	Medical	1	7	1	Male	40	3	1	Laboratory Technician	2	Married	3468	16632	9	Y	No	12	3	4	80	1	6	3	3	2	2	2	2
5	32	No	Travel_Frequently	1005	Research & Development	2	2	Life Sciences	1	8	4	Male	79	3	1	Laboratory Technician	4	Single	3068	11864	0	Y	No	13	3	3	80	0	8	2	2	7	7	3	6
6	59	No	Travel_Rarely	1324	Research & Development	3	3	Medical	1	10	3	Female	81	4	1	Laboratory Technician	1	Married	2670	9964	4	Y	Yes	20	4	1	80	3	12	3	2	1	0	0	0
7	30	No	Travel_Rarely	1358	Research & Development	24	1	Life Sciences	1	11	4	Male	67	3	1	Laboratory Technician	3	Divorced	2693	13335	1	Y	No	22	4	2	80	1	1	2	3	1	0	0	0
8	38	No	Travel_Frequently	216	Research & Development	23	3	Life Sciences	1	12	4	Male	44	2	3	Manufacturing Director	3	Single	9526	8787	0	Y	No	21	4	2	80	0	10	2	3	9	7	1	8
9	36	No	Travel_Rarely	1299	Research & Development	27	3	Medical	1	13	3	Male	94	3	2	Healthcare Representative	3	Married	5237	16577	6	Y	No	13	3	2	80	2	17	3	2	7	7	7	7